Very Slow IO on High Sierra about vessel HOT 6 CLOSED

One way that helps is adding :cached after all volumes on the docker-compose.yml, you can see more info about it here:

https://stories.amazee.io/docker-on-mac-performance-docker-machine-vs-docker-for-mac-4c64c0afdf99
https://docs.docker.com/docker-for-mac/osxfs-caching/

I saw some good improvements here and no issues so far. I thought about opening a PR and defaulting i to cached (linux and windows basically ignores that flag).

Also, the new version released yesterday has some good performance improvements for High Sierra:

https://docs.docker.com/docker-for-mac/release-notes/#stable-release-notes

For systems running APFS on SSD on High Sierra, use raw format VM disks by default. This improves disk throughput (from 320MiB/sec to 600MiB/sec in dd on a 2015 MacBook Pro) and disk space handling.
Existing disks are kept in qcow format, if you want to switch to raw format you need to “Remove all data” or “Reset to factory defaults”. See https://docs.docker.com/docker-for-mac/faqs/#disk-usage

You go on Docker > Preferences > Remove all data and it will start docker from scratch using the new Docker.raw format.

from vessel.

Yep, that's certainly a thing with Docker on Mac currently - it uses a layer of virtualization since Mac's kernal (unix) doesn't play well with Docker (which needs Linux). Unfortunately I don't have control over that (altho I don't really have requests that slow on my projects- perhaps they're smaller. Perhaps is a MacBook - laptop - vs iMac - desktop - thing with access speeds).

Valet will always be faster on Mac since it's running PHP/Nginx (or Caddy, I forget which) directly on your Mac.

from vessel.

from vessel.

That are some neat findings @dbpolito!

from vessel.

Thanks!

In case anyone else was wondering where to find Remove all data.
Docker > Preferences > Reset (💣in right corner)> Remove all data

from vessel.

We overcame this issue by synchronizing the local and the docker for mac filesystem using syncthing. We built an open source tool that follows this approach, in case it helps:
https://github.com/okteto/cnd

from vessel.